First Colloquium on AI for Scientific Discovery

ICTP is launching a new series of colloquia dedicated to AI for scientific discovery that will explore how advances in artificial intelligence (AI) are transforming the way we approach fundamental scientific questions. The series will bring together leading researchers working at the intersection of AI and science, and highlight emerging methods, breakthroughs, and new opportunities for discovery across disciplines.

In the first colloquium of this series, Sebastian Ehlert, Senior Researcher at Microsoft Research AI, will present research at the intersection of AI and natural sciences, spanning materials discovery, molecular synthesis, biomolecular modelling, and electronic-structure theory. In particular, he will focus on Skala, a deep learning-based exchange-correlation functional that aims to make Density Functional Theory (DFT) systematically more accurate and predictive by learning directly from high-quality quantum-chemical data.

The colloquium will take place at 11:00 hrs CEST on Wednesday, 14 October in ICTP’s Euler-Lagrange Lecture Hall, Leonardo Building.

Sebastian Ehlert is a Senior Researcher at Microsoft Research AI for Science, working on machine-learning approaches to DFT and highly accurate quantum chemistry at scale. His research includes the development of the Skala functional and the Microsoft Research Accurate Chemistry Collection (MSR-ACC). He also contributes to the development of extended tight binding (xTB) methods and actively maintains open-source scientific software.